Out-of-home advertising — billboards, transit ads, airport placements, street furniture, digital OOH (DOOH) screens — has no native mechanism for recording who saw an ad. A billboard doesn't know who drove past it. A digital screen in a mall rotates creative on a schedule and logs that an ad played, but not who looked at it. This puts OOH at the far end of the attribution-difficulty spectrum covered in this track: podcast advertising at least has a listener who heard audio and can act on a promo code; OOH often doesn't even have that, because there's rarely a natural moment for the viewer to take an immediate action.

OOH measurement approaches, by what they actually measure

Method	What it measures	Core limitation
Traffic/circulation counts	Estimated people who passed the location (DOT traffic data, foot traffic studies)	Passing ≠ seeing; no attention or recall data
Mobile location data (geofencing)	Devices that entered a radius around the placement, matched to later conversion/visit	Correlational — can't prove the ad caused the visit, only co-located
Lift studies (geo/matched market)	Conversion or brand-metric difference between exposed vs. unexposed matched markets	Requires real experimental discipline; easy to run badly and get false lift
Brand/recall surveys	Self-reported ad recall and brand association in exposed geography	Self-report bias, small sample sizes, timing-sensitive
Promo codes/vanity URLs on the creative	Direct response action tied to that specific placement	Captures only the motivated minority who act on it

Mobile location data as the modern proxy

The most common contemporary workaround is mobile location data: a measurement vendor draws a geofence around the physical placement, identifies devices (via SDK partnerships or data exchanges, anonymized and aggregated) that entered that geofence during the campaign window, and then tracks a subset of those same devices to see if they later visited a target location — a store, a dealership, a branch — within an attribution window. This gives OOH something that looks like a conversion funnel, but it's an approximation built on device co-location, not ad exposure: a device being near the billboard doesn't confirm the person looked up and read it, and the technique can't be applied at all to placements with no clear "did they later visit a physical location" conversion to track, like a brand-awareness-only campaign with no retail footprint.

ooh_geofence_lift_setup.txt

text

Placement: Billboard, Highway 101 northbound, Exit 23
  Geofence: 500m radius around billboard, exposure window: campaign dates
  Exposed device pool: devices entering geofence during window
  Control pool: matched demographic devices, no geofence entry
  Conversion event: visit to target store location within 14 days
  Output: Visit rate (exposed) vs. visit rate (control) = incremental lift %

Geofence exposure is not ad exposure

A device entering the geofence around a billboard only proves the person's phone was physically near the sign — it says nothing about whether they looked at it, whether it was obstructed, whether it was nighttime and the sign wasn't illuminated, or whether they were a passenger glancing at their phone instead of the road. Vendors selling 'OOH attribution' products are selling a correlation between location and later behavior, dressed in attribution language; treat the resulting lift numbers as directional, not causal, unless paired with a genuine matched-market experimental design.

Lift studies: the more rigorous (and more expensive) alternative

A properly designed geo lift study compares a market where the OOH campaign ran against a matched control market where it didn't, tracking a shared conversion or brand metric (sales, site visits, search volume, survey-measured recall) across both over the same period. Because it's a market-vs-market comparison rather than a device-vs-device one, it avoids some of the co-location fallacy that geofencing data carries, but it requires enough markets and budget to run multiple test/control pairs, careful selection of genuinely comparable markets, and enough campaign duration and spend to produce a detectable signal above normal market noise. This is why lift studies are typically reserved for larger OOH budgets — a single-market, single-placement campaign usually can't support one.

Digital OOH (DOOH) adds impression counts, not person-level data

Programmatic DOOH screens can report how many times a creative played and, via connected traffic data for that time slot, an estimated audience count — a meaningful upgrade over static billboards, which have no play log at all. But this is still an estimated opportunity-to-see count derived from traffic modeling, not a confirmed individual exposure, so DOOH's reporting sophistication shouldn't be mistaken for the kind of person-level measurement available in digital channels.

Promo codes and QR codes still do real work here

As with podcast advertising, a unique promo code or QR code printed on OOH creative gives a direct, unambiguous attribution signal for the subset of viewers who act on it — useful as a supplementary signal and as a floor on true causal impact (since every redemption is unambiguously caused by seeing that specific placement), even though it undercounts the brand-building effect on everyone who was influenced but never used the code.

What's next

Affiliate marketing sits at the opposite end of the measurement spectrum from OOH — every affiliate-driven sale is, in principle, directly trackable — but it introduces its own distinct set of attribution and fraud problems worth understanding before trusting affiliate reporting at face value.

Next: Affiliate Marketing Programs →

I build these systems professionally.

Whether it's a RAG pipeline, analytics migration, or AI workflow — let's talk.

Work With Me →

Need production AI or MarTech expertise? I build custom LLM workflows, server-side GTM tagging, and analytics infrastructures professionally.

Need custom AI or MarTech setup? Let's build together.

Let's talk →